Establishing Pipeline Mechanical Integrity

New technology can certainly help in this area. However, over-reliance on one, tried-and-tested system can lead to defects being missed. This is what happened at Prudhoe Bay. Ultrasound methods said that there was no problem, smart pigging said otherwise. Now both systems are regularly used.

Although Prudhoe Bay appeared to be an Internal Corrosion problem, we continually come across similar scenarios on external coating problems; one system fails to find all coating defects.

This reliance on one type of system has been recognised by the NACE ECDA initiative that now requires at least two complementary systems to survey a pipeline. Although the ECDA programme is specific to the USA, it is also relevant worldwide.

Let’s examine what ‘complementary’ means in the context of identifying coating defects and the threat such defects present to the mechanical integrity of a pipeline.

Even with the introduction of the NACE ECDA Standard RP 0502-2002, coating defect detection is still being implemented with instrumentation that was essentially designed in the 1940’s, even if the circuitry has been brought up to date. There is nothing inherently wrong with these systems, indeed they work very well. However they do have limitations.

These ‘old’ technologies use the same principle; by detecting a signal or CP that has made its way to the surface for detection. CIPS, DCVG, ACVG and Pearson all use this technique and in the final analysis are basically the same system (whether or not they use CP). DCVG, ACVG and Pearson look for voltage gradients on the ground surface, whilst CIPS looks for an absolute value referenced to the pipeline, but still relies on ground surface conduction. These are not really complementary systems, if one system misses a defect, they all will.

Whilst these systems do find coating defects, there are some defects that are not found. Why is this?

This can be divided into two areas:

1. Survey methodology limitations
2. Survey technology limitations

Survey methodology for the systems mentioned, require that every metre of the pipeline is surveyed. This may not occur for a number or reasons, namely inaccessibility, accidentally missing a pipeline section and sections that ‘cannot’ be surveyed (historically, sections that are not surveyed for whatever reason, never actually get surveyed).

Survey technologies for the systems mentioned, as has already been stated, are very similar. The detection of a defect relies upon the defect voltage appearing at the surface of the ground, above the pipeline and above the defect. Anything that prevents this, or prevents the voltage appearing in the correct place, will fail to show a defect.

Masked defects are a well documented effect. This is where the current leaking through a defect, either in the presence of another defect, or failure in finding a surface route, fails to show a voltage at the surface – hence the defect is not found. Any system that relies on CP or switched CP being detected by contact means at the surface is not going to find a masked defect. Similarly any system that was designed to use near-DC signals to simulate CP behaviour will also fall foul of the masking effect.

There is also the ‘needle-in-a-haystack’ scenario, where there are many pinhole defects over a large length (even the entire length) of a pipeline. Finding these defects doesn’t really work as there is no ‘good’ coating to reference to so there is not a pronounced voltage gradient at the surface, hence the general degradation of the coating is missed and no value for the coating integrity is obtained.

The systems mentioned so far rely on measuring a signal that has left the pipeline and relying on it finding its way to the surface. If this cannot be measured or found then a defect cannot be detected.

If a signal can be measured that is on the pipeline (rather than what is in the ground), then, regardless of the exit route it takes and whether or not it reaches the surface a measure of coating quality and defect location can occur.

C-Scan works in precisely this manner, by measuring the magnetic field gradient of a signal applied to the pipeline. If the current is lost to ground through a defect, by whatever route, this shows up as reduced current on the pipeline past the defect. Even if the current through the defect doesn’t make it to the surface, the defect will still be found, as C-Scan is inherently non-contact. Thus masking effects are circumvented, and surveys can be easily made over non-conductive surfaces.

As C-Scan is non-contact and measures the current attenuation over a length of pipeline, traditionally difficult areas to survey, such as crops, livestock, rivers or busy roads, can be surveyed by measuring the attenuation over the inaccessible section. An absolute value of coating integrity can therefore be obtained, which is repeatable, operator independent and recordable. This also ensures that no pipeline section is missed that is inherent with traditional systems. If the diameter and wall thickness of the pipe is known, an estimate of the average coating conductance can be given, in real-time, in the field. In close-interval mode, C-Scan can be used to pinpoint defects after a section of pipeline is found to have suspect coating integrity.

By not using surface voltage gradients, C-Scan is truly complementary to existing methods that do.

However, finding coating damage is only part of the story. The end result of an ECDA programme is, or is intended to be, the establishment of an acceptable level of mechanical integrity for the pipeline. Therefore, the knowledge that coating damage exists, is in itself not enough. To this knowledge must be added a value of probability that corrosion is, or may be, active. It can be suggested that there are three basic levels of concern:

Level One – corrosion is active
Level Two – corrosion may be active
Level Three – corrosion is not occurring.

These three levels are obtained by using coating defect analysis and average coating condition, in conjunction with an analysis of cathodic protection activity (or lack of), at sites of coating failure.

Differentiating between these three levels allows the pipeline operator to prioritise resources to address the situations found.

For new or rehabilitated pipelines, ensuring that the coating is in excellent condition after backfill by running a quick C-Scan survey is an excellent way to ensure fewer problems in the future.

A C-Scan current attenuation survey is an approved method within NACE RP0502-2002 so can legitimately be used. The PHMSA ensures that pipeline companies comply with the CFR. RP0502-2002 is a part of the CFR.

Contact:
Peter R Barnes
Chief Engineer
Dynalog Electronics Ltd
Bristol, UK
www.dynalog.co.uk